This paper goes through some of the challenges and opportunities in using AI to tackle microfluidics problems. The paper suggests that since time lapse experiments can typically generate more than 100GB of data, the ability of researchers to process this data is the real bottleneck and AI would be well suited to tackle this problem. The paper then also goes through some microfluidic specific problems and how they translate to AI solutions.
Source: Riordon, Sovilj, Sanner, Sinton, & Young, Deep learning with microfluidics for biotechnology
